There are a variety of situations in which automated processing of a number of data items is desirable. One specific example of such an application is product catalogs. Product catalogs, whether in the form of a paper catalog or an Internet “Web” site, frequently have numerous pictures which each depict a respective one of the various items that are available for sale. Many years ago, these pictures were prepared using optical negatives and photographs. Currently, however, the trend is to maintain and process these pictures in the form of computer files containing digital images.
A given paper or on-line catalog will usually include products from a variety of different manufacturers, and it is common for each manufacturer to provide its own digital images. There will typically be variation between the form of images provided by different manufacturers, for example in terms of characteristics such as the size, shape, resolution, tint, and so forth. It is even possible that the images from a single given manufacturer may have different forms. Accordingly, in order for the images throughout a catalog to have a generally similar appearance, the various images from various sources need to be processed to adjust characteristics such as size, shape, resolution, and/or tint, so as to bring them into general conformity with each other.
A further consideration is that a manufacturer's images do not represent a static situation, because manufacturers are constantly adding new products with new images, discontinuing existing products and associated images, and providing updated images for existing products. Moreover, there may be other reasons for adjusting images. For example, with respect to a paper or on-line catalog intended for use during the Christmas season, there may be a desire to put a festive frame around each image, such as a frame of holly leaves and berries. Moreover, stylistic changes in the images are often desirable.
The traditional approach for carrying out these various types of image processing tasks has involved manual adjustments effected on an image-by-image basis, through use of image processing software requiring extensive operator interaction. However, this is extremely time consuming and expensive. Many organizations currently employ a number of graphic artists to do this work, at great expense.
A less common approach has been the preparation of a hard-coded software routine to process images, written in line-by-line source code. However, these routines are time-consuming and expensive to generate, are likely to include errors or “bugs”, and have little flexibility because they cannot be modified quickly and cheaply. Moreover, they can only be prepared and executed by a skilled programmer, rather than by a graphic artist who is skilled in image processing but has limited computer skills. It is difficult to find persons who have both artistic and computer skills, and they command large salaries.
Thus, while these traditional approaches have been generally adequate for their intended purposes, they have not been satisfactory in all respects. In this regard, a further consideration is that, to optimize efficiency, a computer program which effects automated processing of multiple items of data may be executed for a relatively long period of time, for example several weeks or several months. Computer programs may contain subtle bugs which do not present significant problems where the program is restarted rather frequently, for example on a daily basis, but which can produce relatively significant problems where the program is executed continuously over a long period of time. The problems which result can manifest themselves in various ways, ranging from operational characteristics that become progressively less efficient, to a complete “crash” of the software in which no meaningful processing takes place.